INDEX

Explanations

information related to giving and receiving tests or assessments

oai_token-act-pair · gpt-3.5-turbo

references to educational assessments and testing contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Canaver

-0.65

 Rewards

-0.62

utsu

-0.59

chemy

-0.59

cipled

-0.59

etsk

-0.58

 isEnabled

-0.58

tarian

-0.57

Â¶

-0.56

stellar

-0.56

POSITIVE LOGITS

 bombed

0.63

 passers

0.63

 ninety

0.60

 Ethiop

0.58

aturdays

0.58

 indistinguishable

0.57

".

0.57

 chanting

0.56

ixty

0.54

 thirty

0.53

Activations Density 2.280%

information related to giving and receiving tests or assessments

references to educational assessments and testing contexts

No Comments

No Known Activations